GRS decode_to_code and decode_to_message fail on messages in the code

[cprior]

Currently, the following occurs

sage: F = GF(59)
sage: length = 40
sage: dimension = 12
sage: eval_pts = F.list()[:length]
sage: col_mults = F.list()[1:length+1]
sage: C = codes.GeneralizedReedSolomonCode(eval_pts, dimension, col_mults)
sage: msg = random_vector(F, dimension)
sage: c = C.encode(msg)
sage: c in C
True
sage: c.decode_to_code(c)
7 # random; first entry of msg (expected c)
sage: c.decode_to_message(c)
Traceback (most recent call last)
...
TypeError: 'sage.rings.finite_rings.integer_mod.IntegerMod_int' object is not iterable

We expect:

...
sage: c.decode_to_code(c) == c
True
sage: c.decode_to_message(c) == msg
True

[jlavauzelle]

Hi,

I think you mixed codeword and code in your script. Both decode_to_code and decode_to_message are methods of the code's class. So the good syntax is:

sage: C.decode_to_code(c) == c
sage: C.decode_to_message(c) == msg

Julien

[jlavauzelle]

Hi again,

My mistake, I didn't run the most recent version of Sage. Your typo fooled me, but you're completely right, there's indeed an issue with these methods.

Julien

[dlucas]

Hello,

Just a few remarks:

• when you push a fix on a ticket, don't forget to change its state to needs_review
• you can also put your (full) name on the Authors field
• and when it's a bug fix, you can add a new doctest which shows the bug has been fixed, e.g.

EXAMPLES:

The bug detailed in trac #20340 has been fixed::

    sage: C = codes.GeneralizedReedSolomonCode(GF(59).list()[:40], 12)
    sage: c = C.random_element()
    sage: D = C.decoder("BerlekampWelch")
    sage: D.decode_to_code(c) == c
    True

Otherwise, it seems all good to me!

Best,

David

---

New commits:

​4bca107

Fixed error on _decode_to_code_and_message() methods

[jlavauzelle]

Thanks David,

Following your advice, I added the doctest and some short summaries of the decoders in the top documentation of the grs.py file.

Best,

Julien

[git]

Branch pushed to git repo; I updated commit sha1. New commits:

​e4bdd3d

Add a doctest and fix documentation.

[dlucas]

Hello,

I copied your doctest in every method which was previously impacted by the bug. If you agree with my changes, set it to positive_review, on my side, I'm fine with your fix.

David

---

New commits:

​d4bbfdd

Replicated doctest for the fixed bug in every (once) flawed method

[jlavauzelle]

Nice, thanks a lot,

Julien